Impact of the dynamic vegetation on climate extremes during the wheat growing period over China

Extreme temperature and precipitation indices have important implications for the crop growing season. Whether a coupled regional model with carbon-nitrogen cycling (CN) and vegetation dynamics (DV) can better represent these indices during the growing season compared with a model without these modules remains unknown. This study evaluates the performance of extreme indices in three wheat planting regions (including northeast spring wheat, north winter wheat and south winter wheat regions) over China in the period of 1990-2009 using the Regional Climate Model (RegCM) coupled with the Community Land Model (CLM), which include CN and DV. The results show that relative to the RegCM-CLM, both the RegCM-CLM-CN and RegCM-CLM-CN-DV perform better in simulating summer days (SU), consecutive dry days (CDD), consecutive wet days (CWD), and the interannual variability in all the extreme indices in the three regions but produce larger biases on frost days (FD). The trends of extreme indices in the high-impact risk region of wheat are also better captured by the RegCM-CLM with CN or CN-DV compared with the model without these modules. In the northeast spring wheat and southern winter wheat regions, the greater cold bias of mean daily minimum temperature between RegCM-CLM-CN-DV and RegCM-CLM is consistent with the leaf area index (LAI) difference, which may increase evaporative cooling and thus increasing FD biases. Overestimation of the LAI may have a weaker effect than the surface albedo on the mean daily maximum temperature, leading to decreased SU biases in RegCM-CLM-CN-DV relative to RegCM-CLM.